There is no universal threshold to say that a difference (e.g., 8.5 to 8.7) is automatically “significant.” Beyond sample size, interpretation also depends on:
the usual stability of the indicator (some scores change very little from one year to the next),
the dispersion of responses,
and, above all, the organizational context.
In employee surveys, statistical interpretation is often combined with a “field” perspective:
Have concrete initiatives been implemented since the last survey?
Were these actions targeted at this specific issue?
Do the observed changes align with what teams actually experienced?
A modest difference can therefore be highly meaningful if it is consistent with the actions taken and qualitative feedback. Conversely, a larger difference should be interpreted with caution if no significant changes occurred.
In short, the value of a change lies as much in the context and actions taken as it does in the numbers themselves.